- Email: [email protected]
Antibiotic consumption and Enterobacteriaceae skin colonization in hospitalized adults
Accepted Manuscript Antibiotic consumption and Enterobacteriaceae skin colonization in hospitalised adults Andrew Kirby, Claire Berry, Robert West PII:
S0195-6701(16)30400-5
DOI:
10.1016/j.jhin.2016.09.012
Reference:
YJHIN 4923
To appear in:
Journal of Hospital Infection
Received Date: 11 July 2016 Accepted Date: 13 September 2016
Please cite this article as: Kirby A, Berry C, West R, Antibiotic consumption and Enterobacteriaceae skin colonization in hospitalised adults, Journal of Hospital Infection (2016), doi: 10.1016/j.jhin.2016.09.012. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Title: Antibiotic consumption and Enterobacteriaceae skin colonization in hospitalised adults Running title: Enterobacteriaceae skin colonization Article type: Short report
RI PT
Author names: Andrew Kirby1, 2*, Claire Berry1, Robert West3 Author affiliations 1
SC
Old Medical School, Leeds General Infirmary, Leeds Teaching Hospitals NHS Trust, UK
2
M AN U
Old Medical School, Leeds General Infirmary, University of Leeds, UK
3
Charles Thackrah Building, University of Leeds, UK
*
Corresponding author. Tel: +44 113 233 9239/ +44 7787 26089; E-mail: [email protected]
Summary
TE D
Institution where work was done: University of Leeds
Enterobacteriaceae are increasingly antibiotic resistant and skin colonization may contribute to their
EP
spread in hospitals. We screened one hundred hospitalised adults for Enterobacteriaceae skin colonization and assessed potential risk factors, including antibiotic consumption. Multivariable
AC C
analysis identified antibiotic consumption whilst an inpatient (odds ratio 3.16, 95% confidence interval 1.19 to 8.4) and being male (2.92, 1.06-8.4) were risk factors for Enterobacteriaceae skin colonization. If these risk factors are confirmed, work to understand the biological mechanism involved may lead to the development of interventions to prevent Enterobacteriaceae skin colonization. Keywords: Antibiotic, Colonization, Dermal, Enterobacteriaceae, Skin
1
ACCEPTED MANUSCRIPT Introduction Enterobacteriaceae, including Escherichia coli and Klebsiella pneumoniae, are pathogens associated with significant mortality in hospitalised patients. Evidence suggests antibiotic treated hospitalised patients are at high risk of skin colonization with Enterobacteriaceae. One study dating from 1972
RI PT
determined that 70% of hospitalised patients treated with antibiotics had Klebsiella skin colonization. This compared to 0% in those not treated with antibiotics.1 Skin colonization is a potentially
important route for the transmission of Enterobacteriaceae.2 Transmission may occur by patient to
SC
patient spread and/or via the hands of healthcare workers. Skin colonization with
Enterobacteriaceae may therefore contribute to outbreaks of antibiotic resistant Enterobacteriaceae
M AN U
in hospitals. Data implicating antibiotic use with Enterobacteriaceae skin colonization are old and it is uncertain how relevant the data remains. In view of the increasing healthcare threat posed by antibiotic resistant Enterobacteriaceae to hospitalised patients, we undertook a study of Enterobacteriaceae skin colonization to determine how common it is and whether antibiotic use is a
Methods
TE D
risk factor.
EP
Study population: The study was a prospective cohort feasibility study. Adult patients admitted to medical and surgical wards at Leeds Teaching Hospitals NHS Trust with capacity to consent were
AC C
eligible to participate; those consuming antibiotics on admission to hospital or within one month prior to hospital admission were excluded. Participating patients were sampled at single time points only i.e. this was not a longitudinal study. Clinical data were collected on factors which were considered likely to impact on Enterobacteriaceae skin colonization including antibiotic use, see Table I. Sample collection: Bacteria were isolated from patient’s skin using dry cotton swabs which were transported in the manufacturer’s transport media (Sterilin, M40 Transystem), between 9 am and 5
2
ACCEPTED MANUSCRIPT pm; no restrictions on personal hygiene were required prior to sampling. The following sites were sampled: axilla, buttock, forearm-hand, groin and nose (intranasal). The duration of the swabbing procedure was not defined in the study protocol. Ethical approval was obtained from the Health Research Authority (158402) and patients provided informed consent. Samples were collected
RI PT
specifically for this research project. Sampling method: The principal research question was whether hospital inpatients consuming antibiotics were at increased risk of Enterobacteriaceae skin colonization. Therefore participants
SC
were assigned to one of two groups (antibiotic group or no antibiotic group) based on their inpatient antibiotic use. The antibiotic group included patients who had consumed two or more doses of
M AN U
antibiotic whilst an inpatient by the time of sampling. The non-antibiotic group, for pragmatic reasons, had consumed ≤ 1 dose of antibiotic at the time of sampling (1 dose = ≥1 antibiotic given at one time point). A risk of biasing the antibiotic group to older patients, patients who had been in hospital for longer and patients from certain clinical specialties was identified. To limit this potential
TE D
bias, patients in the antibiotic and non-antibiotic groups were matched where possible by duration of hospitalisation (stratified to the following groups; <24 hours, 1-7 days and 8-28 days), clinical
EP
speciality (medical and surgical) and by age (+/- 15 years) (See Table II). Bacterial identification: Swabs were inoculated onto blood and cysteine lactose electrolyte deficient
AC C
(CLED) agar (E&O, Bonnybridge, UK) and incubated at 37°C for 24 hours. Bacteria were identified using a matrix assisted laser desorption ionisation time of light (MALDI-TOF) mass spectrometry (Bruker, Germany). Antimicrobial susceptibility testing of Enterobacteriaceae was completed according to EUCAST standards. Skin colonization was defined as isolation of ≥1 colony of a bacterial species on an agar plate. Sample size: It was anticipated the rate of Enterobacteriaceae skin colonization was at least 20%. Since the primary analysis was logistic regression, a sample size of 100, giving 20 colonization events,
3
ACCEPTED MANUSCRIPT was determined to provide robust estimates for the coefficients of two risk factors allowing the risk of antibiotics to be assessed.3 Statistical analysis: Statistical analyses were performed using Microsoft Open R version 3.2.3. Comparisons of variables of interest between those patients with/without Enterobacteriaceae skin
RI PT
colonization used Pearson’s Chi-squared test. Continuous variables were first categorised to aid interpretation and account for nonlinearities. All variables of interest were considered in a logistic regression of Enterobacteriaceae skin colonization. Every combination of eight main effects was
SC
considered, giving 256 models. The model with the lowest value of Akaike’s Information Criterion (using R package glmulti version 1.0.7) was selected and reported. 4 The final model was assessed
M AN U
using a Receiver Operating Characteristic (ROC) curve reporting the Area Under the Curve (AUC), sensitivity and specificity, which quantify its prediction performance. The ROC curve was generated using the Epi package version 1.1.71.
TE D
Results
Patient demographics: In total 100 inpatients were recruited during 2015, see Table I. They included sixty patients having received ≤1 dose of antibiotic and forty patients having received >2 doses of
EP
antibiotic. The characteristics of these groups, including rate of Enterobacteriaceae colonization, with regard to antibiotic use is reported in Table I.
AC C
Risk factors for Enterobacteriaceae skin colonization: The Enterobacteriaceae skin colonization rate in patients having received >2 doses of antibiotic was 38% (15/40), higher than the 15% (9/60) in those receiving ≤1 dose of antibiotics (p=0.012). Univariable analysis identified male sex and inpatient antibiotic use as risk factors for Enterobacteriaceae skin colonization. These were confirmed after multivariable analysis. The adjusted odds ratio for antibiotic use was 3.16 (95% confidence interval, 1.19 to 8.4) and for male sex was 2.92 (1.06 to 8.4). In an assessment of the model, as a result of plotting the ROC curve for the fitted logistic regression, the AUC was
4
ACCEPTED MANUSCRIPT determined as 0.71, and the optimum cut point gave sensitivity of 45.8% and specificity of 84.2%. Given the simplicity of the predictor, based only on male sex and antibiotic consumption, the performance of the model was satisfactory. Only non-antibiotic patients were included in the patients sampled within 24 hours of their hospitalisation, therefore duration of hospitalisation was
RI PT
assessed as a risk factor for Enterobacteriaceae skin colonization. In the exploration of 256 models, no significant association between duration of hospitalisation and Enterobacteriaceae skin
colonization was noted. Nevertheless, the analysis was repeated without those patients who were
SC
sampled within 24 hours of their admission to hospital and similar results obtained, although with wider confidence intervals reflecting the lower sample size. The odds ratio for risk of
M AN U
Enterobacteriaceae skin colonization was 3.22 (95% confidence interval 1.06, 9.73) for male sex, and 2.97 (0.98, 9.02) for inpatient antibiotic consumption.
Site of Enterobacteriaceae colonization: Colonization by Enterobacteriaceae was found to mainly occur in the groin, with 13% of patients colonized, 7% were colonized in the axilla, 5% in the nose,
TE D
5% on the buttock and 1% on the forearm and hand.
Bacterial species and antibiotic use: The Enterobacteriaceae species identified included three
EP
Citrobacter, seven E.coli, four Enterobacter, seven Klebsiella, one Morganella, three Pantoea (formally Enterobacter) and four Proteus. The antibiotic susceptibility of these isolates was unrelated
AC C
to patient’s antibiotic use. In total 26 different antibiotic regimens were administered to the 40 patients in the antibiotic group. Discussion
Hospitalisation has been cited as a risk factor for colonization with Enterobacteriaceae, but evidence supporting this may have been biased by the lack of matching between inpatients and outpatients.5, 6
Our data suggests that hospitalisation is not a risk factor for skin colonization with
Enterobacteriaceae, with patients in hospital for <24 hours having the same rate of skin colonisation
5
ACCEPTED MANUSCRIPT as those hospitalised for 1-28 days who did not consume antibiotics. Our data suggest that antibiotic consumption and being male are risk factors for skin colonization with Enterobacteriaceae in hospitalised adults. This data is relatively novel as the most studies of Enterobacteriaceae colonization have focused almost exclusively on gastrointestinal colonization with resistant
RI PT
Enterobacteriaceae.7 Most Enterobacteriaceae skin colonization occurs in the groin and axilla, and it has been suggested before that the groin may act as a permanent source of Enterobacteriaceae skin colonization in
SC
some patients.8 Both the groin and axilla are rich in sweat glands and this may explain why these sites are most frequently colonized, and why males, with more active sweat glands, are more
M AN U
commonly colonized.9 The finding that colonization was higher for patients on antibiotics may relate to the effect of antibiotics on the skin microbiome. Alterations (loss) to the skin microbiome may reduce the resistance to colonization normally provided by commensal bacteria, so called colonization resistance. Interactions between commensal bacteria and the ability of Staphylococcus
colonization resistance.10
TE D
aureus to colonise the nasal cavity have been shown and provide evidence to support the concept of
EP
The study was undertaken in a clinical setting and was pragmatically designed. Furthermore, the patients were heterogeneous with regard to their underlying conditions and antibiotic regimens.
AC C
Despite these challenges we identified potential risk factors for Enterobacteriaceae skin colonization with a relatively small number of patients. This small size may have prevented the identification of risk factors which are less strongly associated with Enterobacteriaceae skin colonization. In addition there may have been risk factors which have not been identified and hence not accounted for in the multivariable analysis. Conclusions
6
ACCEPTED MANUSCRIPT Health care staff should be aware that hospitalised male patients and antibiotic treated patients may be at increased risk of Enterobacteriaceae skin colonisation. It may be appropriate in Enterobacteriaceae outbreaks to apply additional infection prevention and control interventions to these patients to reduce Enterobacteriaceae transmission.
Katharine Pelton and Dr Jane Turton, Public Health England.
SC
Conflicts on interest: None
RI PT
Acknowledgements: Research Nurses Claire Brown, Ajie Jaguraga, Faye Pinker, Elizabeth Smith and
Source of Funding: This work was supported by a grant from Leeds Teaching Hospitals Charitable
M AN U
Funding. Neither the funder nor sponsor had any role in study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the article for
AC C
EP
TE D
publication.
7
ACCEPTED MANUSCRIPT
RI PT
References
1- Pollack M, Charache P, Nieman RE, Jett MP, Reimhardt JA, Hardy PH Jr.. Factors influencing colonization and antibiotic-resistance patterns of gram-negative bacteria in hospital patients
SC
Lancet 1972; 30: 668-71.
1977; 2: 1315–1317.
M AN U
2- Casewell, M. W, Phillips I. Hands as a route of transmission for Klebsiella species Br Med J
3- Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis J Clin Epidemiol 1996; 49: 1373-9.
19: 716–723.
TE D
4- Akaike H. A new look at the statistical model identification IEEE Trans Automat Contr 1974;
5- Larson EL, McGinley KJ, Foglia AR, Talbot GH, Leyden JJ. Composition and antimicrobic resistance of skin flora in hospitalized and healthy adults J Clin Microbio. 1986; 23: 604-8.
EP
6- Larson EL, Cronquist AB, Whittier S, Lai L, Lyle CT, Della Latta P. Differences in skin flora between inpatients and chronically ill outpatients. Heart Lung 2000; 29:298-305.
AC C
7- Tischendorf J, de Avila RA, Safdar N. Risk of infection following colonization with carbapenem-resistant Enterobactericeae: A systematic review Am J Infect Control 2016; S0196-6553: 01235-3.
8- Lin MY, Lolans K, Blom DW, Lyles RD, et al; Centers for Disease Control and Prevention Epicenter Program. The effectiveness of routine daily chlorhexidine gluconate bathing in reducing Klebsiella pneumoniae carbapenemase-producing Enterobacteriaceae skin burden among long-term acute care hospital patients Infect Control Hosp Epidemiol 2014; 35: 440-2.
8
ACCEPTED MANUSCRIPT 9- Madeira LG, da Fonseca MA, Fonseca IA, de Oliveira KP, Passos RL, Machado-Moreira CA, Rodrigues LO. Sex-related differences in sweat gland cholinergic sensitivity exist irrespective of differences in aerobic capacity Eur J Appl Physiol 2010; 109: 93-100. 10- Sollid JU, Furberg AS, Hanssen AM, Johannessen M. Staphylococcus aureus: determinants of
RI PT
human carriage. Infect Genet Evol 2014; 21: 531-41. Table I: Demographics and Enterobacteriaceae skin colonization rates for study participants Inpatient No antibiotics 40 63 16 (40%) 29
<24 hours No antibiotics 20 56 11 (55%) 28
84 (84%)
30 (75%)
34 (85%)
20 (50%)
100%
40 (100%)
40 (100%)
20 (100%)
35 (35%) 10 (10%) 1 (1%) 24 (24%)
20 (50%) 1 (3%) 1 (3%) 15 (38%)
14 (35%) 0 (0%) 0 (0%) 6 (15%)
1 (5%) 9 (45%) 0 (0%) 3 (15%)
M AN U
SC
Inpatient Antibiotics 40 61 23 (56%) 31
AC C
EP
TE D
Number of participants Age: years (mean) Sex: Male BMI (mean/range) Level of dependence Self caring Usual residence Own home Antibiotics consumed within 3 months before admission Use of pre-admission chlorhexidine Use of inpatient chlorhexidine Enterobacteriaceae skin colonization
All participants 100 61 50 (50%) 29.6
9
ACCEPTED MANUSCRIPT
patients within each stratification group. Length of stay
Antibiotics
05
05
No-antibiotics
10
10
8-28 days Medical Surgical patients patients
SC
1-7 days Medical Surgical patients patients
M AN U
<24 hours Medical Surgical patients patients
RI PT
Table II: Stratification and matching of the patients sampled in the study, with the total number of
101
102
103
104
101
102
103
104
AC C
EP
TE D
1-4: Matched by age. 5: No patients included as it was not practically possible to sample patients who had >2 doses of antibiotics within the first 24 hours of their hospital stay.
10
S0195-6701(16)30400-5
DOI:
10.1016/j.jhin.2016.09.012
Reference:
YJHIN 4923
To appear in:
Journal of Hospital Infection
Received Date: 11 July 2016 Accepted Date: 13 September 2016
Please cite this article as: Kirby A, Berry C, West R, Antibiotic consumption and Enterobacteriaceae skin colonization in hospitalised adults, Journal of Hospital Infection (2016), doi: 10.1016/j.jhin.2016.09.012. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Title: Antibiotic consumption and Enterobacteriaceae skin colonization in hospitalised adults Running title: Enterobacteriaceae skin colonization Article type: Short report
RI PT
Author names: Andrew Kirby1, 2*, Claire Berry1, Robert West3 Author affiliations 1
SC
Old Medical School, Leeds General Infirmary, Leeds Teaching Hospitals NHS Trust, UK
2
M AN U
Old Medical School, Leeds General Infirmary, University of Leeds, UK
3
Charles Thackrah Building, University of Leeds, UK
*
Corresponding author. Tel: +44 113 233 9239/ +44 7787 26089; E-mail: [email protected]
Summary
TE D
Institution where work was done: University of Leeds
Enterobacteriaceae are increasingly antibiotic resistant and skin colonization may contribute to their
EP
spread in hospitals. We screened one hundred hospitalised adults for Enterobacteriaceae skin colonization and assessed potential risk factors, including antibiotic consumption. Multivariable
AC C
analysis identified antibiotic consumption whilst an inpatient (odds ratio 3.16, 95% confidence interval 1.19 to 8.4) and being male (2.92, 1.06-8.4) were risk factors for Enterobacteriaceae skin colonization. If these risk factors are confirmed, work to understand the biological mechanism involved may lead to the development of interventions to prevent Enterobacteriaceae skin colonization. Keywords: Antibiotic, Colonization, Dermal, Enterobacteriaceae, Skin
1
ACCEPTED MANUSCRIPT Introduction Enterobacteriaceae, including Escherichia coli and Klebsiella pneumoniae, are pathogens associated with significant mortality in hospitalised patients. Evidence suggests antibiotic treated hospitalised patients are at high risk of skin colonization with Enterobacteriaceae. One study dating from 1972
RI PT
determined that 70% of hospitalised patients treated with antibiotics had Klebsiella skin colonization. This compared to 0% in those not treated with antibiotics.1 Skin colonization is a potentially
important route for the transmission of Enterobacteriaceae.2 Transmission may occur by patient to
SC
patient spread and/or via the hands of healthcare workers. Skin colonization with
Enterobacteriaceae may therefore contribute to outbreaks of antibiotic resistant Enterobacteriaceae
M AN U
in hospitals. Data implicating antibiotic use with Enterobacteriaceae skin colonization are old and it is uncertain how relevant the data remains. In view of the increasing healthcare threat posed by antibiotic resistant Enterobacteriaceae to hospitalised patients, we undertook a study of Enterobacteriaceae skin colonization to determine how common it is and whether antibiotic use is a
Methods
TE D
risk factor.
EP
Study population: The study was a prospective cohort feasibility study. Adult patients admitted to medical and surgical wards at Leeds Teaching Hospitals NHS Trust with capacity to consent were
AC C
eligible to participate; those consuming antibiotics on admission to hospital or within one month prior to hospital admission were excluded. Participating patients were sampled at single time points only i.e. this was not a longitudinal study. Clinical data were collected on factors which were considered likely to impact on Enterobacteriaceae skin colonization including antibiotic use, see Table I. Sample collection: Bacteria were isolated from patient’s skin using dry cotton swabs which were transported in the manufacturer’s transport media (Sterilin, M40 Transystem), between 9 am and 5
2
ACCEPTED MANUSCRIPT pm; no restrictions on personal hygiene were required prior to sampling. The following sites were sampled: axilla, buttock, forearm-hand, groin and nose (intranasal). The duration of the swabbing procedure was not defined in the study protocol. Ethical approval was obtained from the Health Research Authority (158402) and patients provided informed consent. Samples were collected
RI PT
specifically for this research project. Sampling method: The principal research question was whether hospital inpatients consuming antibiotics were at increased risk of Enterobacteriaceae skin colonization. Therefore participants
SC
were assigned to one of two groups (antibiotic group or no antibiotic group) based on their inpatient antibiotic use. The antibiotic group included patients who had consumed two or more doses of
M AN U
antibiotic whilst an inpatient by the time of sampling. The non-antibiotic group, for pragmatic reasons, had consumed ≤ 1 dose of antibiotic at the time of sampling (1 dose = ≥1 antibiotic given at one time point). A risk of biasing the antibiotic group to older patients, patients who had been in hospital for longer and patients from certain clinical specialties was identified. To limit this potential
TE D
bias, patients in the antibiotic and non-antibiotic groups were matched where possible by duration of hospitalisation (stratified to the following groups; <24 hours, 1-7 days and 8-28 days), clinical
EP
speciality (medical and surgical) and by age (+/- 15 years) (See Table II). Bacterial identification: Swabs were inoculated onto blood and cysteine lactose electrolyte deficient
AC C
(CLED) agar (E&O, Bonnybridge, UK) and incubated at 37°C for 24 hours. Bacteria were identified using a matrix assisted laser desorption ionisation time of light (MALDI-TOF) mass spectrometry (Bruker, Germany). Antimicrobial susceptibility testing of Enterobacteriaceae was completed according to EUCAST standards. Skin colonization was defined as isolation of ≥1 colony of a bacterial species on an agar plate. Sample size: It was anticipated the rate of Enterobacteriaceae skin colonization was at least 20%. Since the primary analysis was logistic regression, a sample size of 100, giving 20 colonization events,
3
ACCEPTED MANUSCRIPT was determined to provide robust estimates for the coefficients of two risk factors allowing the risk of antibiotics to be assessed.3 Statistical analysis: Statistical analyses were performed using Microsoft Open R version 3.2.3. Comparisons of variables of interest between those patients with/without Enterobacteriaceae skin
RI PT
colonization used Pearson’s Chi-squared test. Continuous variables were first categorised to aid interpretation and account for nonlinearities. All variables of interest were considered in a logistic regression of Enterobacteriaceae skin colonization. Every combination of eight main effects was
SC
considered, giving 256 models. The model with the lowest value of Akaike’s Information Criterion (using R package glmulti version 1.0.7) was selected and reported. 4 The final model was assessed
M AN U
using a Receiver Operating Characteristic (ROC) curve reporting the Area Under the Curve (AUC), sensitivity and specificity, which quantify its prediction performance. The ROC curve was generated using the Epi package version 1.1.71.
TE D
Results
Patient demographics: In total 100 inpatients were recruited during 2015, see Table I. They included sixty patients having received ≤1 dose of antibiotic and forty patients having received >2 doses of
EP
antibiotic. The characteristics of these groups, including rate of Enterobacteriaceae colonization, with regard to antibiotic use is reported in Table I.
AC C
Risk factors for Enterobacteriaceae skin colonization: The Enterobacteriaceae skin colonization rate in patients having received >2 doses of antibiotic was 38% (15/40), higher than the 15% (9/60) in those receiving ≤1 dose of antibiotics (p=0.012). Univariable analysis identified male sex and inpatient antibiotic use as risk factors for Enterobacteriaceae skin colonization. These were confirmed after multivariable analysis. The adjusted odds ratio for antibiotic use was 3.16 (95% confidence interval, 1.19 to 8.4) and for male sex was 2.92 (1.06 to 8.4). In an assessment of the model, as a result of plotting the ROC curve for the fitted logistic regression, the AUC was
4
ACCEPTED MANUSCRIPT determined as 0.71, and the optimum cut point gave sensitivity of 45.8% and specificity of 84.2%. Given the simplicity of the predictor, based only on male sex and antibiotic consumption, the performance of the model was satisfactory. Only non-antibiotic patients were included in the patients sampled within 24 hours of their hospitalisation, therefore duration of hospitalisation was
RI PT
assessed as a risk factor for Enterobacteriaceae skin colonization. In the exploration of 256 models, no significant association between duration of hospitalisation and Enterobacteriaceae skin
colonization was noted. Nevertheless, the analysis was repeated without those patients who were
SC
sampled within 24 hours of their admission to hospital and similar results obtained, although with wider confidence intervals reflecting the lower sample size. The odds ratio for risk of
M AN U
Enterobacteriaceae skin colonization was 3.22 (95% confidence interval 1.06, 9.73) for male sex, and 2.97 (0.98, 9.02) for inpatient antibiotic consumption.
Site of Enterobacteriaceae colonization: Colonization by Enterobacteriaceae was found to mainly occur in the groin, with 13% of patients colonized, 7% were colonized in the axilla, 5% in the nose,
TE D
5% on the buttock and 1% on the forearm and hand.
Bacterial species and antibiotic use: The Enterobacteriaceae species identified included three
EP
Citrobacter, seven E.coli, four Enterobacter, seven Klebsiella, one Morganella, three Pantoea (formally Enterobacter) and four Proteus. The antibiotic susceptibility of these isolates was unrelated
AC C
to patient’s antibiotic use. In total 26 different antibiotic regimens were administered to the 40 patients in the antibiotic group. Discussion
Hospitalisation has been cited as a risk factor for colonization with Enterobacteriaceae, but evidence supporting this may have been biased by the lack of matching between inpatients and outpatients.5, 6
Our data suggests that hospitalisation is not a risk factor for skin colonization with
Enterobacteriaceae, with patients in hospital for <24 hours having the same rate of skin colonisation
5
ACCEPTED MANUSCRIPT as those hospitalised for 1-28 days who did not consume antibiotics. Our data suggest that antibiotic consumption and being male are risk factors for skin colonization with Enterobacteriaceae in hospitalised adults. This data is relatively novel as the most studies of Enterobacteriaceae colonization have focused almost exclusively on gastrointestinal colonization with resistant
RI PT
Enterobacteriaceae.7 Most Enterobacteriaceae skin colonization occurs in the groin and axilla, and it has been suggested before that the groin may act as a permanent source of Enterobacteriaceae skin colonization in
SC
some patients.8 Both the groin and axilla are rich in sweat glands and this may explain why these sites are most frequently colonized, and why males, with more active sweat glands, are more
M AN U
commonly colonized.9 The finding that colonization was higher for patients on antibiotics may relate to the effect of antibiotics on the skin microbiome. Alterations (loss) to the skin microbiome may reduce the resistance to colonization normally provided by commensal bacteria, so called colonization resistance. Interactions between commensal bacteria and the ability of Staphylococcus
colonization resistance.10
TE D
aureus to colonise the nasal cavity have been shown and provide evidence to support the concept of
EP
The study was undertaken in a clinical setting and was pragmatically designed. Furthermore, the patients were heterogeneous with regard to their underlying conditions and antibiotic regimens.
AC C
Despite these challenges we identified potential risk factors for Enterobacteriaceae skin colonization with a relatively small number of patients. This small size may have prevented the identification of risk factors which are less strongly associated with Enterobacteriaceae skin colonization. In addition there may have been risk factors which have not been identified and hence not accounted for in the multivariable analysis. Conclusions
6
ACCEPTED MANUSCRIPT Health care staff should be aware that hospitalised male patients and antibiotic treated patients may be at increased risk of Enterobacteriaceae skin colonisation. It may be appropriate in Enterobacteriaceae outbreaks to apply additional infection prevention and control interventions to these patients to reduce Enterobacteriaceae transmission.
Katharine Pelton and Dr Jane Turton, Public Health England.
SC
Conflicts on interest: None
RI PT
Acknowledgements: Research Nurses Claire Brown, Ajie Jaguraga, Faye Pinker, Elizabeth Smith and
Source of Funding: This work was supported by a grant from Leeds Teaching Hospitals Charitable
M AN U
Funding. Neither the funder nor sponsor had any role in study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the article for
AC C
EP
TE D
publication.
7
ACCEPTED MANUSCRIPT
RI PT
References
1- Pollack M, Charache P, Nieman RE, Jett MP, Reimhardt JA, Hardy PH Jr.. Factors influencing colonization and antibiotic-resistance patterns of gram-negative bacteria in hospital patients
SC
Lancet 1972; 30: 668-71.
1977; 2: 1315–1317.
M AN U
2- Casewell, M. W, Phillips I. Hands as a route of transmission for Klebsiella species Br Med J
3- Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis J Clin Epidemiol 1996; 49: 1373-9.
19: 716–723.
TE D
4- Akaike H. A new look at the statistical model identification IEEE Trans Automat Contr 1974;
5- Larson EL, McGinley KJ, Foglia AR, Talbot GH, Leyden JJ. Composition and antimicrobic resistance of skin flora in hospitalized and healthy adults J Clin Microbio. 1986; 23: 604-8.
EP
6- Larson EL, Cronquist AB, Whittier S, Lai L, Lyle CT, Della Latta P. Differences in skin flora between inpatients and chronically ill outpatients. Heart Lung 2000; 29:298-305.
AC C
7- Tischendorf J, de Avila RA, Safdar N. Risk of infection following colonization with carbapenem-resistant Enterobactericeae: A systematic review Am J Infect Control 2016; S0196-6553: 01235-3.
8- Lin MY, Lolans K, Blom DW, Lyles RD, et al; Centers for Disease Control and Prevention Epicenter Program. The effectiveness of routine daily chlorhexidine gluconate bathing in reducing Klebsiella pneumoniae carbapenemase-producing Enterobacteriaceae skin burden among long-term acute care hospital patients Infect Control Hosp Epidemiol 2014; 35: 440-2.
8
ACCEPTED MANUSCRIPT 9- Madeira LG, da Fonseca MA, Fonseca IA, de Oliveira KP, Passos RL, Machado-Moreira CA, Rodrigues LO. Sex-related differences in sweat gland cholinergic sensitivity exist irrespective of differences in aerobic capacity Eur J Appl Physiol 2010; 109: 93-100. 10- Sollid JU, Furberg AS, Hanssen AM, Johannessen M. Staphylococcus aureus: determinants of
RI PT
human carriage. Infect Genet Evol 2014; 21: 531-41. Table I: Demographics and Enterobacteriaceae skin colonization rates for study participants Inpatient No antibiotics 40 63 16 (40%) 29
<24 hours No antibiotics 20 56 11 (55%) 28
84 (84%)
30 (75%)
34 (85%)
20 (50%)
100%
40 (100%)
40 (100%)
20 (100%)
35 (35%) 10 (10%) 1 (1%) 24 (24%)
20 (50%) 1 (3%) 1 (3%) 15 (38%)
14 (35%) 0 (0%) 0 (0%) 6 (15%)
1 (5%) 9 (45%) 0 (0%) 3 (15%)
M AN U
SC
Inpatient Antibiotics 40 61 23 (56%) 31
AC C
EP
TE D
Number of participants Age: years (mean) Sex: Male BMI (mean/range) Level of dependence Self caring Usual residence Own home Antibiotics consumed within 3 months before admission Use of pre-admission chlorhexidine Use of inpatient chlorhexidine Enterobacteriaceae skin colonization
All participants 100 61 50 (50%) 29.6
9
ACCEPTED MANUSCRIPT
patients within each stratification group. Length of stay
Antibiotics
05
05
No-antibiotics
10
10
8-28 days Medical Surgical patients patients
SC
1-7 days Medical Surgical patients patients
M AN U
<24 hours Medical Surgical patients patients
RI PT
Table II: Stratification and matching of the patients sampled in the study, with the total number of
101
102
103
104
101
102
103
104
AC C
EP
TE D
1-4: Matched by age. 5: No patients included as it was not practically possible to sample patients who had >2 doses of antibiotics within the first 24 hours of their hospital stay.
10